AutoPerfPy UI Report

Performance Analysis Report

Collector: PsutilCollectorRun Label: cpu_0_fp32_bs2Data Source: tests/fixtures/tool_outputs/autoperfpy_real_output.jsonDevice: Intel64 Family 6 Model 141 Stepping 1, GenuineIntelAccelerator: cpu_0Precision: fp32Batch Size: 2CPU: Intel64 Family 6 Model 141 Stepping 1, GenuineIntelOS: Windows 11Streams: 1Generated: 2026-02-21 21:22:38
Samples
8
P99 Latency
55.76 ms
Mean Throughput
17.1 FPS
Mean Power
29.1 W
Duration
2.0 s
P95 Latency
54.69 ms

Platform & Run Metadata

Device Intel64 Family 6 Model 141 Stepping 1, GenuineIntel
Accelerator cpu_0
CPU Intel64 Family 6 Model 141 Stepping 1, GenuineIntel
Precision fp32
Batch 2
OS Windows 11

Executive Summary

Key performance metrics at a glance

Samples
8
P99 Latency
55.76ms
P95 Latency
54.69ms
P50 Latency
42.62ms
Mean Latency
46.82ms
Mean Throughput
17.1FPS
Min Throughput
12.7FPS
Mean Power
29.1W
Max Power
56.7W
Avg GPU
0.0%
Avg CPU
48.8%
Mean Memory
15560MB
Max Temp
47.7C
Duration
2.0s

Latency

Latency analysis and distribution

Latency Over Time

Latency Distribution

Memory

Memory usage analysis

Memory Over Time

Memory Usage

Power & Thermal

Power consumption and temperature monitoring

Power Over Time

Temperature Over Time

Throughput

Throughput performance analysis

Throughput Over Time

Utilization

CPU and GPU resource utilization

CPU/GPU Utilization Over Time

Utilization Summary

Raw Data

Preview of collected samples used to compute summary metrics and charts.

Sample Preview (first 8 rows)

elapsed_secondslatency_mscpu_percentgpu_percentmemory_used_mbmemory_percentpower_wtemperature_cthroughput_fpsis_warmup
021.92930015512.652396.356.6647.6917.5887true
0.607760.232241.9015538.566496.420358.203true
0.809453.881335.1015532.16896.420358.4553true
1.010841.907626.4015549.644596.5203510.5124true
1.212359.928834.3015586.730596.720359.9343true
1.414841.801636.8015616.730596.9203519.9883false
1.616642.623845.3015565.20796.6203518.5947false
1.819656.027264.3015498.7596.2203512.6541false

Result Guide

Plain-English purpose of each key result and how to interpret it for decisions.

Metric Purpose Guide

MetricWhat It MeansWhy It Matters
sample_countTotal collected measurement points in this run.Higher sample counts increase confidence in regression and trend decisions.
warmup_samplesInitial samples excluded from steady-state analysis.Prevents startup transients from distorting production-latency conclusions.
duration_secondsTotal measured run time in seconds.Validates benchmark window length and comparability between runs.
latency.mean_msAverage latency across measured samples.Useful trend metric, but can hide tail outliers.
latency.p50_msMedian response latency.Represents typical user experience and baseline responsiveness.
latency.p95_ms95th percentile latency.Captures tail latency for near-worst-case requests.
latency.p99_ms99th percentile latency.Primary straggler signal; spikes indicate queueing, throttling, or contention.
throughput.mean_fpsAverage processed samples per second.Primary capacity metric for planning and SLO sizing.
throughput.min_fpsLowest observed throughput in steady state.Highlights instability or periodic slowdowns under load.
cpu.mean_percentAverage CPU utilization.Identifies host bottlenecks, preprocessing pressure, or dataloader saturation.
gpu.mean_percentAverage GPU utilization.Low utilization with high latency usually indicates input or sync bottlenecks.
memory.mean_mbAverage memory footprint in MB.Tracks memory pressure and headroom for larger batch sizes/models.
power.mean_wAverage power draw in watts.Used for energy-cost modeling and efficiency tracking.
power.max_wPeak power draw in watts.Helps detect power spikes and validate power envelope constraints.
temperature.max_cMaximum observed device temperature.High peaks correlate with thermal throttling risk and performance collapse.

Run Metadata

Platform, configuration, and run metadata captured for this benchmark.

Run Overview

FieldValue
CollectorPsutilCollector
Run Labelcpu_0_fp32_bs2
Sample Count8
Warmup Samples5
Duration (s)2.0221

Platform Metadata

FieldValue
device_nameIntel64 Family 6 Model 141 Stepping 1, GenuineIntel
device_idcpu_0
device_typeCPU
soc-
gpu_model-
cpu_modelIntel64 Family 6 Model 141 Stepping 1, GenuineIntel
power_mode-
osWindows 11

Inference Configuration

FieldValue
precisionfp32
batch_size2
acceleratorcpu_0
streams1
warmup_runs5
iterations100

Summary Details

Expanded summary metrics generated from the run payload.

Detailed Summary Metrics

MetricValueUnit
Sample Count8
Warmup Samples5
Duration (s)2.0221
latency.mean_ms46.8175ms
latency.min_ms41.8016ms
latency.max_ms56.0272ms
latency.p50_ms42.6238ms
latency.p95_ms54.6869ms
latency.p99_ms55.7591ms
throughput.mean_fps17.079FPS
throughput.min_fps12.6541FPS
throughput.max_fps19.9883FPS
power.mean_w29.0711W
power.peak_w64.98W
power.max_w56.66W
temperature.mean_c36.5862C
temperature.max_c47.69C
cpu.mean_percent48.8%
cpu.max_percent64.3%
cpu.min_percent36.8%
gpu.mean_percent0%
gpu.max_percent0%
memory.mean_mb15560.2292MB
memory.max_mb15616.7305MB
memory.min_mb15498.75MB
memory.total_mb16110.7852MB
memory.mean_percent96.5667%

Report Information

Report metadata and generation details

Generated2026-02-21 21:22:38
AuthorTrackIQ
Total Figures0
Data Sourcetests/fixtures/tool_outputs/autoperfpy_real_output.json
CollectorPsutilCollector
Run Labelcpu_0_fp32_bs2
DeviceIntel64 Family 6 Model 141 Stepping 1, GenuineIntel
Precisionfp32
Batch Size2
CPUIntel64 Family 6 Model 141 Stepping 1, GenuineIntel
OSWindows 11
Acceleratorcpu_0
Streams1
Warmup Runs5
Iterations100
Samples8
Duration (s)2.0221